Open Arabic LLM - Arabic MMLU General Knowledge (Middle School): leaderboard
The "Arabic MMLU General Knowledge (Middle School)" subtask of Open Arabic LLM. A task from the Open Arabic LLM Leaderboard, evaluating models on Arabic-language knowledge, reasoning and understanding across native and translated benchmarks.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 163 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen 2.5 72B Instruct | 81.4 | #454 |
| 2 | gemma-3-27B-pt | 77.91 | #502 |
| 3 | SILMA-9B-Instruct-v1.0 | 77.91 | #913 |
| 4 | Qwen 2.5 32B Instruct | 76.74 | #498 |
| 5 | Qwen2.5-32B-Instruct-abliterated-v2 | 76.74 | #394 |
| 6 | Qwentile2.5-32B-Instruct | 76.74 | #339 |
| 7 | Awqward2.5-32B-Instruct | 76.74 | #371 |
| 8 | Saka-14B | 76.16 | #344 |
| 9 | Qwen2.5-32B-Instruct-CFT | 76.16 | #375 |
| 10 | Qwen 2.5 72B | 75.58 | #499 |
| 11 | Llama 3.3 70B Instruct | 75 | #534 |
| 12 | free-evo-qwen72B-v0.8-re | 75 | #435 |
| 13 | Gemma 3 27B (IT) | 74.42 | #523 |
| 14 | Qwen 2.5 14B | 74.42 | #725 |
| 15 | AIC-1 | 74.42 | #231 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-general-knowledge-middle-school · How It Works · Data refreshed daily, snapshot 2026-10-20.